A Unique Leukemia Mouse Model Established From AML Patient with IDH2 R140Q and FLT3-ITD Mutations Among Other Common AML Mutations

Abstract 3579

Somatic mutations in isocitrate dehydrogenase 1 (IDH1, cytosolic) and IDH2 (mitochondrial homolog) have been associated with acute myeloblast leukemia (AML, 7% and 15% respectively), as well as other malignancies, e.g. gliomas, etc^1,2. These mutations lead to 2-hydroxyglutarate (2HG) production in leukemic cells, which can be detected both in leukemic cells and plasma of AML patients. This association has led to speculation that IDH mutations are important in AML leukemogenesis, or oncogenic drivers, and could potentially be used as drug targets for AML treatment. However, these assumptions have never been verified for lack of experimental models. We recently successfully engrafted leukemic cells from bone marrow of an AML patient into immunocompromised NOD/SCID mice. This patient has aggressive disease of poor prognosis, with initial response to standard of care of chemotherapy and subsequent fast recurrence, followed by death. The patient was diagnosed as M5 subtype AML with heterozygous IDH2 R140Q mutation, and also mutations of FLT3-ITD, DNMT3A R882H and NPM1. The transplanted mice (AM7577) developed AML leukemia with typical symptoms (BW loss, hunched, inactivity, labored breathing, ruffled coat and eventual mortality) and with leukemic cells in bone and peripheral organs (e.g. spleen, blood, etc) (gradual detection of leukemic cells in peripheral blood after 30 days). The leukemic cells can serially be passed in mice with 100% take-rate and cause 100% leukemia induced mortality (even with < 1e5 cells), thus creating a renewable and potentially unlimited source of leukemia cells. The leukemic cells in mice are identical to those of the original patient leukemic cells (CD45⁺, CD33⁺, CD13⁺, CD123⁺, and CD19⁻, heterozygous IDH2 R140Q mutation, along with FLT3-ITD; DNMT3A R882H and NPM1. This mutation pattern is somewhat in contrast to the previous belief that IDH mutations are mutually exclusive to other common AML mutation³. The engrafted animals produce 2HG in leukemic cells and in mouse plasma, both of which can be readily detected and quantified by high-performance liquid chromatography and mass spectrometry (HPLC/MS). 2HG and cytokines could be useful biomarkers to monitor the disease progression as well as treatment response. The treatment leukemic models in vivo with standard of care (SOC: Ara-C, 5day-on/2day-off dosing scheme) resulted in significant response, manifested by reduction of leukemic load to undetectable in peripheral blood and serum biomarkers for both tested doses (2 and 10mg/kg), as well as extended survivals. Leukemia, however, rapidly relapse after the treatment withdrawal, suggesting that AraC has effect only on the leukemic blasts, but not on leukemic stem cells. The appearance of leukemic blasts in peripheral blood after araC withdrawal can be effectively re-suppressed by araC. Further tests on Flt3 inhibitor for anti-leukemic activities are currently ongoing. In summary, the engrafted mice could serve as a useful experimental model to investigate the role of IDH mutation in leukemogenesis and help to identify new treatment strategies for AML, including compounds targeting IDH mutations and FLT3, etc. This is, to our knowledge, the first reported AML experimental model of featuring IDH mutation.